Admittedly, I still believe that constraints on various supply levels—primarily electricity, not just generation capacity, but actual supply access for the latest data centers—will bring some uncertainty to capital expenditure this year and next.
It will also, to a large extent, cause massive fluctuations in sentiment. Whether temporary optimism or pessimism, both are risks.
However, simple market fluctuations have never been the subject of my discussion here, have they?
Nearly thirty years ago during a summer break, I wrote a set of DOS-based drawing programs using Pascal and Assembly. One of my classmates wrote a "Tank Battle" game in C, and another wrote "Monopoly" in BASIC...
A year later, in the third year of junior high, due to my lagging Chinese grades, my homeroom teacher deemed programming a "distraction from proper business." She believed I should save that time to "patch my weaknesses." Of course, the academic result was good. However, from then on, "writing code" gave me a sense of guilt. The time saved didn't actually help me fix my weaknesses; it even instilled a subconscious fear of the Chinese language. The so-called saved time was purely wasted in an ineffective drift.
Of course, looking back, these things didn't cause me any real loss. The dividends of the era allowed our generation to fare reasonably well as long as we "followed the rules." To this day, I maintain a good relationship with that teacher. I respect her because she didn't force me; she used what was considered an enlightened approach for that time, doing her best to hope I would become better. Seeing photos of her 80th birthday celebration at my alma mater recently, she still looks vigorous. I wish her health and longevity.
Nowadays, no one would consider a junior high student "writing code" to be a distraction from proper business. The times have long since changed.
Another deep impression from my school days was a board-poster competition in high school. It wasn't on a blackboard but eventually landed on paper. I whimsically used Photoshop and Word. Luckily, my homeroom teacher at the time was a pioneer of digital teaching even within the province. But when he helped me print the poster I thought was great, it was less than 16K in size, pitifully squeezed among the A3 or even A3+ layouts of other students. My self-confidence took a hit.
Today, hand-drawing has instead become an absolute scarcity.
Sometimes, I ask myself: if I had persisted in being "unconventional" back then, would I have turned out differently? Would I have had the chance to look up at my childhood idol, Bill Gates, from a slightly smaller angle?
But my cognition at the time wouldn't let me know that it was the start of a new era, the establishment of a "new order."
My personality also caused me to "miss out" on the eras of the Internet, mobile Internet, and Big Data...
But overall, I have lived quite well so far, thanks to the era. I am also grateful for the continuous small setbacks in life, which prevented me from lingering too long on any single technology. They allowed me to try many different hardware and software architectures, systems, programming languages, tools, and more importantly, a wide variety of projects and ideas.
Being a generalist prevented me from becoming too obsessed, allowing me to follow the market beta safely through various challenges.
It is always after an era has passed that one sighs: "Ah, what a great opportunity; if only I had seized it." Then, when a new era arrives, one continues to start high and end low...
In maybe five years, or perhaps two or three, no one will write code themselves, and no one will hand-craft PPTs. The vast majority of content we see will be produced by AI, just like those "board posters" of the past.
The moment Vibe Coding appeared, it became impossible for us to return to the old production workflows, even if models stop improving, and even if the "AGI" we desire is delayed.
Vibe Coding is just the most typical example of the past year. What about others? Search? Design? Knowledge learning? Healthcare? Scientific research?
Perhaps what is unfolding in Davos, Switzerland, right now gives us more confidence to say: the old rules are gone forever.
Of course, not the rules of grand narratives, but rules concerning specific "production," or rather, information-related production.
Compared to a specific individual, AI output is metaphysical, following large-sample laws, and soulless. But from the perspective of a "factory," it's a completely different story. Current models lack sufficient attention, cannot yet think truly like humans, lack memory and self-evolution capabilities, and have many hallucinations. However, as a whole compared to a collective of humans, they are: faster, better, cheaper, and more abundant.
In today's society, perhaps what we criticize has never been "pre-prepared meals," but rather those that are expensive, unpalatable, and deceptively sold. Enjoying fresh-cut and stir-fried food with true "wok hei" will become a "luxury choice" for those with money and time. Cheap, standardized "pre-prepared meals" are the collective choice for the vast majority.
AI is the same.
For a while, I tried to view the "absurdity" of AI-based production processes replacing humans from a "human" perspective. But technology never stands still. None of us has the right to worry for others, let alone for the next generation.
If, when I was a teenager, "writing code" wasn't seen as a distraction, or if someone could have told me then that Photoshop + Word was the future, my life path might have been entirely different—maybe better, maybe worse. But at least I could have been closer to the "creation" I desired.
The life of each generation should not be defined by the worries and concerns of their elders. Because every elder is themselves a product of their own era.
Therefore, we shouldn't worry that junior programmers won't grow because of massive Vibe Coding; we shouldn't worry that young doctors won't get practice because of "AI diagnosis"...
The moment I was born, there was electricity in my home. Although frequent power outages taught me the necessity of keeping candles ready, I can never imagine the state of "studying hard under a kerosene lamp." I can use a coal stove, but I can't start a wood fire. I have skills my father's generation doesn't, even though I will never learn the things their era taught them...
Yes, even until recently, I belonged to the group that "worried about what young people would do in the AI era." But thanks to my somewhat rebellious nature and some views on child education in a famous person's blog, I suddenly realized the dread of being an "old fogey."
That view is to strictly control or even ban children from accessing the Internet and social networks.
Gen-Z is a generation that grew up in an environment of the Internet and social networks. This is precisely not the environment we grew up in. Thus, it is dangerous and irresponsible to simply draw such conclusions based on a sense of "innate superiority" and a "natural critical" stance.
Humanity has never slowed the pace of technological progress because of so-called dissenting voices.
In fact, children's ability and efficiency in obtaining and processing information from the Internet and social media today are things we could not even imagine when we were young. Suppose we were given the chance to be thirty years younger; we would likely rarely complain about the Internet and social networks.
Perhaps the history of human progress is just like this. It's just that we turned from "teenagers" into "old fogeys," and then adopted the same preachy tone we once loathed in our youth.